1. Field of the Invention
The present invention relates to an image processing apparatus and method thereof, which allow usage, modification, display, movement, sharing, and the like of document images.
2. Description of the Related Art
Multi-functional peripheral equipments (MFPs) have gained remarkably advanced functions, and those which have scanner and facsimile functions in addition to copy and print function have become the mainstream in the market. An MFP in an office has an important role as an interface between information printed on paper sheets and digital data.
Especially, the scanner function of the MFP can efficiently transform a volume of documents into digital data. For this reason, documents are scanned and are saved in a database (DB) on a network or of the MFP itself, and undergo document management in many cases. Digital data obtained by scanning a document will be referred to as a “document image” hereinafter.
Along with the popularization of personal computers (PCs), document creation using application software on such PC is common. A document in a digital data state created by the application software will be referred to as a “digital document” hereinafter.
Document images and digital documents have advantages of easy re-editing and re-use of data, easy sharing of data by the large number of persons, no deterioration along with the elapse of time, and the like compared to documents printed on paper sheets. On the other hand, documents printed on paper sheets have advantages of high readability, easy handling, high portability, high understandability, and the like compared to the document images and digital documents. For this reason, in many examples, it is often efficient to print digital documents and document images on paper sheets.
Conversely, document images and digital documents which are the same as or similar to documents printed on paper sheets are often required. For example, the user damages some document pages printed on paper sheets and wants to print the digital document in the initial state. Also, the user wants to re-print a digital document in color, which is originally a color document but was printed as a monochrome document.
There may be many situations in which a photo, text, table, and line art included together in a document image are respectively analyzed, and the analysis results are to be re-used. For example, data transformation based on character recognition when a document image is used as text data, transformation into a vector image free from any deterioration of image quality if a specific region is enlarged to be easy to see, and the like are made in such situations.
In this way, a document management system which has documents printed on paper sheets as a start point and allows the user to easily make use, modify, display, move, share, and like of data is demanded.
When a document image generated by scanning a document that includes a photo and text together is re-printed, for example, an A4-size, JPEG-compressed color scan image of 300 dpi has a data size close to 3.5 MB, and if the large number of document images flow on the network, the load on the network becomes heavy. However, when the resolution is decreased, the image quality of a text region deteriorates. Hence, uniform image processing to be applied to one page of a document image has a limitation, and the following prior arts have been proposed.
Japanese Patent Laid-Open No. 8-204950 discloses image management which segments an input image into a photo region and text region, and sets the resolutions of the photo region and text region to be variable. However, only a raster image has a variable resolution, and a vector image and the character recognition result of the text region are not managed. Japanese Patent Laid-Open No. 2003-348326 discloses a technique for segmenting a document image for respective lines, and re-combining the segmented lines so as to display a scan image on a portable terminal (PDA). However, the output destination is limited to the portable terminal, an object to be processed is only a raster image, and a vector image and character recognition result are not supported.
Japanese Patent Laid-Open No. 2000-306103 discloses a technique for segmenting a document image into a text region, image region, and background region, distributing these regions to layers prepared in advance to greatly reduce the data size of the document image, and holding document data in a format that allows easy re-use of data. Even in document data in a format that allows easy re-use, information of each region is only one type.
A feature common to these prior arts is that images are managed for respective attributes of regions. However, even regions with an identical attribute have various natures. For example, given two different text regions in a single document image, one region may have a high character recognition ratio, but the other region may have not so high a recognition ratio. In this way, even when identical types of information are provided because they have the same attribute, it is hard to say that satisfactory information provision can be made.
These prior arts execute processing such as resolution transformation and the like in a limited way upon data registration, and do not support processing for transforming into a resolution that the user wants upon outputting.
Under the present circumstances, when the user wants to re-print, as a color document, a digital document which is originally a color digital document and was printed as a monochrome document, or when the user wants to acquire a digital document of an original copy based on a document printed on paper sheets, he or she normally inquires a person who created an original digital document to acquire a required digital document.
A wide variety of apparatuses such as printers, monitors, PDAs, and the like are used to output document images, and the output contents may often be a full or partial document image. Furthermore, data which can be modified is often required, and various types of information are required for a document image.
However, in Japanese Patent Laid-Open No. 8-204950, only a raster image can have a variable resolution. A vector image, the character recognition result of a text region, and the like are not managed. Also, in Japanese Patent Laid-Open No. 2003-348326, its output destination is limited to the PDA, the object to be modified is only a raster image, and a vector image and character recognition result are not supported.
There may be many situations in which a photo, text, table, and line art included together on a document are respectively analyzed, and a document is to be output by applying image processing in correspondence with the attributes of objects. The technique in Japanese Patent Laid-Open No. 2000-306103 holds a document image in a format that allows easy re-use, as described above. However, only one type of information of each region is output to another apparatus or application.
These regions manage images for respective attributes of regions. However, as described above, regions have various natures although they have an identical attribute. Even when identical types of information are provided because they have an identical attribute, information cannot often be satisfactorily transferred.
Furthermore, for respective objects of a document, a document cannot be output by replacing data formats or data themselves depending on the output mode of the document, the contents of objects, and the like.